Race-neutral lung function equations alter disease classification, disability compensation

Key takeaways:

• The degree and direction of change in various outcomes with race-neutral vs. race-specific equations differed based on race/ethnicity.
• Discriminative accuracy did not differ between the two equations.

SAN DIEGO — Millions of individuals had different clinical, occupational and financial outcomes with race-neutral vs. -specific lung function equations, according to a presentation at the American Thoracic Society International Conference.

Further, outcomes often changed to a greater degree among Black patients compared with white patients and Hispanic patients, according to researchers.

“In 2023, the American Thoracic Society issued a statement where they argued that pulmonary function testing [PFT] laboratories should adopt a race-neutral approach to PFT interpretation,” Arjun (Raj) Manrai, PhD, assistant professor of biomedical informatics in the Blavatnik Institute at Harvard Medical School, said during the presentation.

“In the same report, they said at the same time that the consequences for the yet unquantified number of individuals with results near decision-making thresholds, around which results are expected to shift after removing race from reference equations, need to be carefully tracked and considered,” Manrai continued.

With access to data from five cohorts, Manrai, James A. Diao, MPhil, MD, who was a fourth-year medical student at Harvard Medical School at the time of the study, and colleagues analyzed spirometry values for 249 million individuals to determine how lung function equations not adjusted for race (race-neutral; Global Lung Function Initiative [GLI]-Global) change clinical, financial and occupational outcomes observed when adjusted for race (race-specific; GLI-2012).

Clinically, researchers found that race-neutral vs. race-specific equations for interpreting lung function meant millions of individuals could have different ventilatory impairment, obstructive or nonobstructive (12.5 million), and COPD severity grades (2.05 million).

For nonobstructive ventilatory impairment specifically, use of race-neutral vs. race-specific equations led to 141% more Black patients with this impairment. In contrast, a reduction in the prevalence of this impairment was found when evaluating white patients (69% reduction) and Hispanic patients (77% reduction).

The same pattern was observed during the assessment of changes in the prevalence of moderate to severe COPD with race-neutral vs. race-specific equations, with a 44% increase among Black individuals, a 15% decrease among white individuals and a 21% decrease among Hispanic individuals.

Researchers considered firefighter eligibility criteria to assess the impact of the two equations on occupational eligibility, and 2.28 million individuals would have a change in eligibility with use of race-neutral vs. race-specific equations, Diao said.

The prevalence of occupational disqualification went up for Black individuals with use of the race-neutral equation (86% increase), whereas this went down for white (–28%) and Hispanic (–43%) individuals.

Using the American Medical Association ratings for medical impairment, Diao highlighted that not adjusting for race vs. adjusting for race resulted in 8.16 million people facing a change in this classification. There was a 120% increase in the prevalence of moderate to severe impairment among Black individuals when comparing the race-neutral equation vs. the race-specific equation. Contrastingly, prevalence decreased by 27% among white individuals and by 18% among Hispanic individuals.

In terms of Veterans Affairs disability compensation, 413,000 individuals experienced a change in receipt of this compensation based on use of a race-neutral equation over a race-specific equation.

Splitting up the cohort based on race/ethnicity revealed a 17.1% higher mean annual payment ($1.1 billion) per Black veteran from Veterans Affairs with a race-neutral equation. When comparing the change in this payment between the two equations among white veterans, researchers did not observe a large change (–1.15%; –$0.52 billion).

In addition to impacting financial outcomes, not adjusting for race meant some patients on the U.S. lung transplant waitlist in 2020 (n = 1,399) could have been bumped up.

“We found the person who would be most advantaged would be a Black woman aged 72 with an FVC of just under 2 liters,” Diao said. “She would experience a decrease in her percent predicted [FVC] of about 10%, which would in fact increase her allocation score, move her up 150 positions on the waiting list and decrease her expected wait time by 4 and a half weeks.”

On the other hand, researchers noted that a white man aged 61 years with an FVC of 2.78 L would have the least advantage when a race-neutral vs. race-specific equation was used.

Instead of moving up on the wait list, Diao noted that this individual would go back 80 spots and wait time would increase by 2.5 weeks.

Notably, discriminative accuracy did not differ between the two equations when researchers evaluated prediction of several outcomes: medical visits for wheezing in the previous year, overnight hospitalization in the previous year, new-onset asthma, new-onset COPD, respiratory-related mortality, all-cause mortality and mortality on the transplant waiting list.

“Despite having no difference for the prediction of longitudinal respiratory outcomes, we’re still seeing the reclassifications of millions of individuals and billions of dollars,” Diao said during the presentation.

“We hope that our data can prompt reconsideration of how these [decision-making] thresholds are drawn and how we’re thinking about assessment of lung health and disease as we move away from these crude heuristics like race,” Diao added.

References:

• Diao JA, et al. N Eng J Med. 2024;doi:10.1056/NEJMsa2311809.
• Measuring lung function more accurately and more equitably. https://hms.harvard.edu/news/measuring-lung-function-more-accurately-more-equitably. Published May 19, 2024. Accessed May 19, 2024.